Creating Markov models in a game of tennis using Maple

The Markov model proposed in the study showed a game duration in tennis of 3.27 minutes consisting of an average of 5.23 points per game. The calculated result differs by 0.275% from a good game time estimate from the US Open 2014 average match times. The winning chances of players were given through a separate smaller 5 × 5 fundamental matrix. The server had a winning chance of 0.926 and the receiver had a winning chance of 0.074. The chances of both players losing with 3 advantages is minuscule with 2.26×10^-3

Download PDF here